Hi guys I am experiencing a problem I have not found any solution for.
Scenario: I have created 3 complex Claygent Prompts that research and qualify energy related metrics for a company. Results are very good so far
Problem: The prompts sometimes take up to 5 minutes each
How can this be lowered significantly?? I want to apply this research large-scale but it takes up to 10 minutes per lead which just won't work for my company.
Here are several effective ways to optimize your Claygent tasks: 1. Break down your complex queries into smaller, more focused batches. Instead of searching for multiple energy metrics at once, split them into separate, targeted searches. 2. Make your search queries more specific and well-defined. This helps Claygent focus on exactly what you need rather than processing unnecessary information. 3. Use our AI Formulas for quick data manipulation and cleaning tasks instead of relying on Claygent for everything. 4. Structure your prompts with: • Clear context at the beginning • Step-by-step instructions • Specific output format requirements 5. Monitor your usage limits and API quotas to ensure you're not hitting any restrictions.For additional help, you can use our metaprompter feature to optimize your prompts for better efficiency. Source: Prompt Engineering Crash Course [Clay 101] - Clay University
Looks like you're still looking for help. We're escalating this over to our support team and someone should be in touch soon. In the meantime, feel free to include any other information that may be helpful in solving your issue!
This is one exemplary prompt, that should provide an estimated energy consumption of the company: Role: You are a research analyst that estimates the yearly energy consumption of companies. Instructions: Based on the provided metrics, estimate the yearly energy consumption of the company. Use the following steps: 1. First Step: ISO 50001 Certification as in ISO 50001 Certification Status. If the company is ISO 50001 certified, assume it consumes at least 1 GWh per year and is likely consuming more than 5 GWh per year, unless there is explicit evidence of lower consumption. 2. Second Step: For each following metric, research industry averages or related insights that are allow for a more realistic estimation of the energy consumption of the company.
Number of employees as in Number Of Employees. You could determine energy consumption per employee for the given industry
Annual revenue as in Annual Revenue. You could estimate energy consumption per revenue unit in the company’s sector.
Industry as in Industry. You could use industry-specific benchmarks (e.g., manufacturing, logistics, healthcare).
Operational size as in Operational Size. You could assess energy use based on the number and type of facilities (e.g., offices, factories, warehouses, clinic beds).
Operations type as in Operations Type. For production-based companies, use output or capacity metrics (e.g., energy per ton produced, per vehicle manufactured).
Reported energy metrics as in Reported Energy Metrics. If CO2 emissions are reported, convert them into approximate energy consumption using standard conversion factors.
3. Third step: Combine Insights and zse all available data and researched benchmarks to derive the most realistic estimate of yearly energy consumption. Output Format: Provide a categorical output of the yearly energy consumption: • <1 GWh • 1-5 GWh • >5 GWh
Hey there Maximilian thanks for reaching out, Do you mind sending the link to the table so we can take a look? https://downloads.intercomcdn.com/i/o/1171717996/ac7729f84b426e1d0358af02/image.png?expires=1732902300&signature=e348462127a716db502ceba07a3d5b3560da1af140ab4cc40d808ba4d14e8e29&req=dSEgF85%2FmohWX%2FMW1HO4zYTywSn1BpUV2BKsam%2FhVPvHbA%3D%3D%0A
Thanks for helping!!
FYI we are thinking about buying the 800$ plan or even the custom solution. The only core issue currently is the processing time.
Hey there Maximillian took a look at this and after some testing on my end was not running into similar time issues, it likely that this may have been due to some latency it was experiencing at the time. A couple of recommendations I do have though is, first off, if you have not already used it yet, you can try using our meta-prompter to help write a more concise version of your prompt. What the prompt will do is take the prompt and re write it into a format that the AI will have an easier time comprehending and executing. The meta prompter can be accessed by clicking on the "Help Me" option in the prompt menu. https://downloads.intercomcdn.com/i/o/w28k1kwz/1270779853/e840a064c7c25e38e4106461d3a8/image.png?expires=1732905000&signature=90ac920de3dede6276347df073a8be0f98fb0ef0991cb4e9262efd9a67746309&req=dSIgFs55lIlaWvMW1HO4zSUXeAoM38g2u%2Fky82lDaeLBwTvfUUtzrhqU%2FPTH%0Ajxf9%0A Secondly, when it comes to more complex prompts such as these even with the meta-prompter help the AI can still take some time for certain prompts. In this case running the column in batches can also help with the process as the AI will be taking in less request at a time. The downside being that you would have to manually run these columns in batches every time as well.
Hi Maximilian J.! This thread was recently closed by our Support team. If you have a moment, please share your feedback:
Thank you so much for sharing your feedback Maximilian J.!
.png){kind=logo}
